Aiding Surveillance

7/25/2019 Aiding Surveillance

1/77


2/77


3/77

03/80

Aiding Surveillance Privacy International

Executive Summary

Introduction

Methodology

Management Information Systems

and electronic transfersDigital identity registration and biometrics

Mobile phones and data

Border surveillance and security

Development at the expense of human rights?The case for caution

Endnotes

Section 1

Section 2

Section 3

Section 4

Section 5

Section 6

Section 7

04

05

15

17

28

42

50

56

59

Contents


4/77

04/80


Executive Summary

Information technology transfer is increasingly a crucialelement of development and humanitarian aid initiatives.Social protection programmes are incorporating digitisedManagement Information Systems and electronic transfers,registration and electoral systems are deploying biometrictechnologies, the proliferation of mobile phones is facilitatingaccess to increased amounts of data, and technologies are

being transferred to support security and rule of law efforts.Many of these programmes and technologies involve thesurveillance of individuals, groups, and entire populations.The collection and use of personal information in thesedevelopment and aid initiatives is without precedent, andsubject to few legal safeguards. In this report we show that asdevelopment and humanitarian donors and agencies rush toadopt new technologies that facilitate surveillance, they maybe creating and supporting systems that pose serious threats

to individuals human rights, particularly their right to privacy.


5/77

05/80

Aiding Surveillance Privacy InternationalSection 1

Introduction

It is hard to imagine a current public policy arena that doesnot incorporate new technologies in some way, whether in theplanning, development, deployment, or evaluation phases.New technologies are enabling the creation of new forms andhigh quantities of data that can inform policy-making processes,improving the effectiveness and efficiency of public policy andadministration. Water management, for example, now employs

measurement and metering techniques; tax administrationincreasingly involves outsourcing of contracts to the privatesector, and the use of data mining techniques for analysis;healthcare now involves advanced diagnostic technologiesand distributed computing.

Today, advanced data analysis technologies and techniquesinform and underpin sustainable policy related to transport,health, infrastructure and other public services. Frequently, thisdata includes vast amounts of personal information of citizens,

and, increasingly, non-citizens. Generating and analysing suchdata creates new and potentially malevolent opportunities forsurveillance the use of personal data to influence, manage,02direct or protect those whose data have been garnered03 bypublic and private entities. As a result, in Europe and NorthAmerica, and increasingly globally, there is a trend towardsthe establishment of legal frameworks to govern how personalinformation is managed and to ensure that individuals rightsare protected.

It is possible to see surveillance as a necessity in modern societies.Over the years leading social thinkers have conceptualisedsurveillance in numerous beneficial ways:04as progress towardsefficient administration, a benefit for the development of Westerncapitalism, essential to the modern nation-state, and even apower generator in itself.05Yet surveillance that is unconstrained

1.0


6/77

06/80


by legal frameworks, human rights protections and the rule oflaw has the potential to jeopardise individuals rights to privacy,free expression, association, assembly and political participation.

As such,in developed countries, the introduction of newtechnologies that have the potential to facilitate surveillancehas traditionally been accompanied by public resistance, critiqueand oversight.

In developing countries, however, new technologies andtechniques are being deployed with a considerably less criticaleye. Analysis of the potential adverse implications of the usepersonal information is often completely neglected in public

administration in developing countries, and governance measuresto ensure protection of personal information are often non-existent.06Emerging economies and developing nations acrossAfrica, Asia and Latin America are seeing the rapid deploymentof technologies that many more developed countries are hesitantto use, such as national identity registries using biometrictechnologies,07and e-health systems 08with national registries ofsensitive personal information, in the absence of legal safeguardsand, indeed, critical analysis. Security techniques, technologies,and programmes are also being transferred to developing

countries prior to the establishment of the necessary democraticand legal safeguards.

The amount of attention devoted to privacy and personalinformation issues in developing countries is growing. This is inpart due to the increased use of technologies by governments andother institutions, but it also reflects the greater use of technologyby citizens. Modern technologies can facilitate the surveillanceof nearly every interaction done by individuals in their homes, on

the streets, as they travel, over telecommunications networks andthe internet. Registration of populations creates a single store ofidentity that can be used for many purposes, including to trackindividuals use of services and interactions with the state andprivate sector in ways previously unforeseen, and unimaginedby the registration systems themselves. Interfering with privacyallows for control to be exerted on individuals, inhibiting their


7/77

07/80


autonomy. In the public sphere this could result in undue attentionto specific individuals and groups, and this focus turns frequentlyto government critics, opposition groups and parties, journalists,

and human rights defenders. Abuse could be less intentional, butequally destructive, where creating data stores allows accidentaldisclosures that place individuals at risk of fraud by malicious thirdparties. Data collected about individuals for one purpose canbe used for other purposes, including monitoring individuals andgroups, creating profiles of their activities, predicting their activities,and discriminating against them.

New technologies hold great potential for the developing world,

and countless development scholars and practitioners have sungthe praise of technology in accelerating development, reducingpoverty, spurring innovation and improving accountability andtransparency.09Indeed, the ICT4D (information and communicationstechnologies for development) movement has come to dominatethe discourse on technology and development, and is at thecentre of discussions about the post-2015 development andhumanitarian agendas. This is, of course, with some good cause new technologies present countless opportunities for expression,connectivity and empowerment. Developing countries lack the

legacy systems and infrastructure long present in the developedworld, and proponents of the deployment of new technologies indevelopment argue that this facilitates the positive leapfroggingeffect. Why should, after all, a developing country deploy paperID cards when it can use biometrics to secure the process ofissuing identity and delivering public services? Similarly, whymanage borders by merely checking people and their possessionswhen we can search through travel histories and other profiles?After all, developing countries also often face a complex

concoction of political instability, rapid population growth, andinequality that raises the stakes when it comes to public servicedelivery or border management.

The problem, as this report will identify, is that there is asystematic failure to critically contemplate the potential ill effectsof deploying technologies in development and humanitarian


8/77

8/80


initiatives, and in turn, to consider the legal and technicalsafeguards required in order to uphold the rights of individualsliving in the developing world.

As privacy rises on the policy agendas of countries across theworld, the contrasting approaches to new technologies in thedeveloped and developing worlds has become increasingly stark.Many of the technologies embraced as being key to effectiveand sustainable development by the development community10have been the subject of extensive debate in advanced Westerndemocracies in recent years. Identity systems and databases thatcollect biometric information including fingerprints, facial scans,

iris information and even DNA, and other expansive registrationsystems have been proposed, resisted, and sometimes rejectedin various countries. In the United Kingdom significant politicalconcern and scrutiny led to the reversal and destruction ofthe National Identity Register and ID card with the Minister incharge of its destruction calling it intrusive and expensive andarticulating concerns about fantastic claims about supposedbenefits.11Israel saw significant debate around its proposedsmart ID and biometric database, with the High Court callinga pilot program extreme and harmful.12The German parliament

decided to deploy next generation passports using biometricsbut explicitly excluded the storing of biometrics on a centralizeddatabase because of privacy concerns. South Koreas policy ofrequiring real-names to access communications was rejectedby the Constitutional Court because it undermined democracy.13National health and genetic databases14and other nationalregisters have been called into question,15data deleted,16and on occasion dismantled17because of privacy and humanrights concerns.

Systems that track individuals at borders and profile movementshave been called into question in Canada,18the US19and Europe,20resulting in systems being abandoned,21and safeguards beingapplied. Recent revelations by US National Security Agencywhistleblower Edward Snowden of extensive and indiscriminatecommunications surveillance systems in the US and abroad have


9/77

9/80


resulted considerable public outcry in the United States, Europeand elsewhere, resulting in action by the European Union,22UnitedNation bodies 23and civil society groups.24

When surveillance technologies are proposed by policy makersin Western democratic states, at least two debates emerge.The first focuses on human rights, civil liberties and the rule of law.The second interrogates the value of a system, its impact, and thecalculable costs against perceived benefits. Social institutions,civil society, regulators, interest groups, government auditors,opposition parliamentarians, scientists and technologists are ableto interrogate each others claimed understanding of the problem,

statements regarding the effectiveness of the technologicalchoices, and whether other solutions are possible with lessercosts. Increasingly, these debates are intertwined. Discussionsaround the US initiative to enhance security of driving licenses,under the REAL ID Act, led to debates over civil liberties andconstitutionality as much as to debates over the size of the costsand the management of the costs.25Similarly, when the NigerianHouse of Representatives recently stalled the procurement of aninternet surveillance system, they did so both for reasons thatit violated constitutional rights and that it breached the Fiscal

Responsibility Act.26

Contrast these debates with the emergence of new technologiesas a key element for delivering development and humanitarianaid in the developing world. The deployment of surveillancetechnologies by development actors, foreign aid donors andhumanitarian organisations is conducted in the complete absenceof any public debate or deliberation. The development discourserarely considers public opinion of the target populations when

approving aid programmes. Even the availability of countervailingperspectives is surprisingly low. Seminal strategy documentslike the UN Office for Humanitarian Affairs Humanitarianismin a Networked Age27or the UN High-Level Panel on thePost-2015 Development Agendas A New Global Partnership:Eradicate Poverty and Transfer Economies through SustainableDevelopment,28pay scant attention to the potential impact of


10/77

10/80


the adoption of new technologies or data analysis techniques onindividuals privacy.

In sum, there are four major problems arising from the increaseduse of development aid to advance surveillance in developingcountries. First, technologies are being deployed that raisesignificant concerns with regards to privacy and other human rights.Second, such technologies may not necessarily be appropriatefor achieving development goals or may have undesirable sideeffects. Third, these technologies are already seen as legally andtechnologically problematic in more developed countries. Fourth,these technologies are deployed in the absence of relevant and

adequate legal frameworks, in contravention of internationalhuman rights and national constitutional requirements. Too oftenthese are the missing dynamics in modern development discoursearound the deployment of technological solutions.

1.1 Human rights and development

Development is not just, or even mostly, about accelerating

economic growth. The core of development is buildingcapacity and infrastructure, bridging historical divisions,ending conflict, addressing social vulnerabilities, andsupporting democratic societies that protect, respect andfulfill human rights.

Traditionally, a chasm existed between the human rights anddevelopment communities, in which the latter group provedgenerally reluctant to engage in debates about international legalobligations and how to reflect the relevant norms in policies at

the domestic and international level.29This has begun to changein recent years, spurred by the call of Kofi Annan to mainstreamhuman rights in all UN agencies in 1997, the expansion of UNhuman rights mechanisms, and by difficulties experienced indevelopment and humanitarian interventions. In a 2005 report,Annan, then Secretary-General, emphasized that the challenges


11/77

11/80


of human rights, development and security are so closelyentwined that none can be tackled effectively in isolation.30

Nevertheless, development and humanitarian aid organisationshave been slow to adopt a rights-based approach to development.It was not until June 2012 that the European Union released anew strategic framework for the administration of foreign aid thatmarried rights and development. In July 2013 USAID for the firsttime elevated human rights to a key objective in its developmentapproach.31The World Bank is under ongoing pressure tomainstream human rights protections in its programmes.32

In a 2012 speech in Senegal on building sustainable partnershipsin Africa, then-US Secretary of State Hillary Clinton spoke tothe important role foreign aid donors plays in promoting rights indevelopment.In comments that were seen as veiled criticism ofother development funding sources,33she contended that fundingmust be carefully deployed:

the United States will stand up for democracy and universalhuman rights, even when it might be easier or moreprofitable to look the other way, to keep the resources

flowing. Not every partner makes that choice, but we doand we will.34

Yet there continues to be a gap between theory and practice,particularly in the application of new technologies in developmentcontexts. The EU Development Fund has supported the issuanceof voter cards and ID cards in Somaliland;35in 2013 USAID putUS$53 million towards a programme that, amongst other things,facilitated the production of national identification cards in Kenya;36

the UK Department for International Development played a keyrole in setting up the M-PESA mobile money system in Kenya incollaboration with Vodafone.37While each of these initiatives hascontributed to development in the respective countries, they havealso raised a number of concerns from a human rights perspectivethat have been all but ignored.


12/77

12/80


1.2 Development and technologies

New technologies are now seen as a crucial element of

development and humanitarian aid initiatives. Indeed, theaid community has often heralded technology as the key toeffectively and efficiently achieving sustainable developmentand overcoming obstacles to delivering humanitarian aid.Technologies have been embraced as a key component of

humanitarianism in the networked age38and will be a priorityfor the post-2015 agenda discussions, constituting one offour thematic focal points at the World Humanitarian Summitin 2015. Technologies are being incorporated into every

development initiative from education to health to elections,and in humanitarian initiatives related to crisis response, fooddelivery and refugee management.

This fervour surrounding ICT4D discourse has been socacophonous as to drown out or, arguably, forestall anycritical analysis of the potential adverse effects of the adoptionof new technologies on human rights and civil liberties. Thisdiscussion paper seeks to fill the gap in of critical research andthinking on this issue.

The paper will focus on critically evaluating four types oftechnologiesor technical modalities applied in the developmentand humanitariansectors: management information systems andelectronic transfers;biometric identification and voter registrationsystems; the use of mobile phones and the data collectedand generated by them; and border surveillance and securitytechnologies. Each of these interventions seeks to create newinformation infrastructures that become national utilities in ways

that require great care and significant scrutiny.

This assessment also provides an opportunity to reflect criticallyupon and reassess policy choices concerning technology. Forinstance, one form of development policy that has been receivinga significant amount of international development funding hasbeen electoral reform and modernisation, often involving the


13/77

13/80


registration of an entire populations biometrics. This has causednumerous problems and challenges, as viewed most recently inKenya. Applying technology in such contexts is challenging, as the

UN Secretary General contended in 2009:

[S]ome of the poorest countries in the world have chosensome of the most expensive electoral processes andtechnology. [] I am concerned about techniques andsystems that might cause a State, in the conduct of itsown elections, to be financially dependent on donors,or technologically dependent on specific vendors forextended periods [E]xperience throughout the world

has shown that it is not the case that the more complexor expensive a system, the more successful the electionswill be.39

The UN Development Programme, which is fundingmuch of theactivity in this domain, responded that they have begun to arguefor technology and electoral processes that are cost-effective,transparent, sustainable, inclusive, accurate, flexible, andsupported by appropriate infrastructure and computer literacy.40Similarly, the independent body that oversees UK aid programmes,

the Independent Commission for Aid Impact, also recognisedthe UN Secretary Generals concerns, and presented somecase studies:

In countries such as Sierra Leone and DRC, the UK hashelped to fund an investment in biometric technology forvoter registration, requiring equipment such as laptops,webcams, fingerprint scanners, colour printers and mobilegenerators. In Malawi, delicate computer equipment used

to collect photo identification of voters was damagedbecause it was transported in the back of uncoveredvehicles. In Sierra Leone, Commissioners saw an ambitiousvoter registration programme being rolled out, usingbiometric data collection technology. Donor and civil societystakeholders noted the risk to the political process of thistechnologically advanced approach. While we acknowledge


14/77

14/80


the potential of new information technologies to strengthenelectoral processes, deploying such sophisticatedtechnologies in difficult environments has a high failure

rate and does not usually represent good value for money.41

While technologies and new programmes may help target, support,and secure development, their adoption must be subjected torights-based questions about whether they are the necessary,proportionate, and effective methods for development, andwhether legal frameworks exist to protect against human rightsabuses. Only after answering these questions can a judgement bemade about whether the right technologies are being deployed in

the most appropriate ways. Importantly, a rights-based evaluationmust come before the critical assessment of the technology; thealternative would allow for an enquiry about the ideal methodsfor deploying problematic technologies.


15/77

15/80


Methodology

This paper draws from the authors expertise and scholarshipin privacy, technology, human rights and development.The authors have been engaged with analysis of technologyin development and humanitarian initiatives since at least2008, and have conducted field research on the issue ofbiometric identification technology in refugee managementsituations. In 2011 we undertook research on medical

information protection in development and humanitarianinitiatives.42The research for this study builds upon this priorresearch and other desk research undertaken over the pastyear, including research undertaken by Privacy Internationalinto privacy in the developing world.

A number of research challenges were encountered whenconducting this review. Surveillance is a domain that is difficult toobserve because it is, by its very nature, secret. Furthermore,researching development programmes is quite challenging due

to the absence of transparency requirements in the design,implementation and evaluation of development programmes.Initiatives such as the International Aid Transparency Initiative andthe Humanitarian Accountability Partnership have gone some wayto alleviate this challenge.

Development initiatives that involve the transfer of technology orcapabilities are also often particularly obfuscated because ofthe involvement of the private sector in providing technologies or

infrastructure essential to the project. While such interventionsoften generate significant interest at the outset, unfortunately thisdoes not translate into a level of transparency across the life-cycleof the programme. Rarely are funding proposals made public.Procurement information is infrequently published. In turn, thespecific types of technologies being sought and delivered cannotbe monitored.

2.0


16/77

16/80


This report thus focuses on only a few international organisations,foreign aid donors, and international funding agencies thatarticulate clearly what it is they are funding at a project-level basis.

Monitoring and evaluation programmes have proven to be quitehelpful in elucidating what the programmes and projects tried toaccomplish and some of the obstacles to success, but even themost critical evaluations have not necessarily critically analysedthe technology in detail, and none identified human rights asa consideration.

A few key evaluations and studies have been relied upon, alongwith reports from foundations and other agencies, when available,

but these are often high-level statements, or selectively detailedpress releases and narratives of the successful achievementsfrom development interventions. One positive trend is the growingnumber of insightful local media organisations and civil societyinstitutions in developing countries that have begun to questionthe merits of technology choices, procurement processes, and thesustainability of development interventions. This paper thereforemakes use of these media reports and perspectives, and althoughthey do represent secondary sources of information, the samecan be said of the published statements from foundations and

international organisations.

When the significant e-government movement in the developedworld expanded in the 1990s and 2000s without critical analysisand at great cost, it took significant critical analysis by academics,media, and civil society to catalyse questioning of the meritsof programmes, technological efficacy, and human rightsimplications.43The same emergence of a policy discourse isbeginning to appear in the developing world. This discourse

certainly needs informing, but many key issues are beingraised already. For instance, there is an emerging sensitivity toprocurement policy, as evidenced by a recent public uproar overtender processes in Kenya,44the Maldives,45and Nigeria.46This paper seizes on, and seeks to further, such discourses.


17/77

17/80


Management Information Systemsand electronic transfers

The promise

In recent years, donors, development agencies and poverty-reduction initiatives have increasingly turned towards socialprotection, cash transfer or social safety net programmesas an effective tool for addressing extreme poverty andaccelerating development in the worlds poorest countries.The term refers to the provision of benefits in cash or in

kind to secure protection in case of social risks and needs,and takes the form of cash transfer schemes, public workprogrammes, social pensions, school stipends and foodvouchers or transfers.47

Although social security systems have played an integral role inmany developed countries for decades, the idea that a minimumlevel of non-contributory social protection could be affordableand easily adopted by low-income countries has really gained

momentum only in the last ten years. Programmes such as BolsaFamilia in Brazil and Oportunidades in Mexico have achievedimpressive advancements in decreasing poverty and improvinghealth and education outcomes. Widespread political supportfor the idea of non-contributory minimum social protectioncrystallised in 2009, when the heads of the UN agencies launchedthe Social Protection Floor Initiative as one of the nine UN

joint initiatives to cope with the global economic and financialcrises. Importantly, the G20 States declared their support forsocial protection in the 2011 Cannes Summit Final Declaration,

emphasising the importance of investing in nationally determinedsocial protection floors which will foster growth resilience, socialjustice and cohesion.48

Social protection is now a priority initiative both for bilateralaid donors, such as the UKs Department for International

3.1


18/77

18/80


Development,49the US Agency for International Development,50and the European Commission 51and for development agencies,such as the World Bank 52and UNICEF.53

Research suggests that social protection initiatives cansignificantly reduce the prevalence and severity of poverty,54contribute to improved nutrition levels, help families absorb thecosts associated with schooling 55and have a positive impacton higher school attendance levels,56reduce child labour,57andimprove maternal health, and the lives of people living with HIV/AIDs. Nevertheless, a number of significant challenges exist whendelivering social protection initiatives in developing countries

which often impede the effectiveness of such programmes.Obstacles include the absence of legal and institutionalframeworks, long-term strategies, and adequate and sustainablefinancing; programme fragmentation and a lack of capacity ofprogramme stakeholders; and institutionalised discriminationand the absence of a gender approach in programme designand implementation. Programmes are also hampered by practicalchallenges associated with, for example, the geographicalremoteness of target communities; difficulties in identifyingpotential beneficiaries; requirements for the production of

identification; transportation, accommodation and opportunitycosts associated with collecting payments in remote or dangerousareas;58and complex application processes which require literacy.

In this context, new technologies are seen to hold enormouspotential and promise for improving the reach and effectivenessof social protection programmes. In recent years, a variety ofICTs have been piloted to increase the reach and effectivenessof social protection programmes, particularly in remote and rural

areas, and include smart cards, cell phones, mobile ATMs, GPSdevices, and biometrics.59In addition, the migration of socialprotection systems from paper-based to fully electronic systemsis gradually being undertaken in many countries, in combinationwith the consolidation of information derived from multiple andseparate social protection initiatives into a Single Registry ofsocial protection beneficiaries. Proponents of the integration of


19/77

19/80


ICTs into social protection programmes cite the following benefits:efficiency and cost-effectiveness; flexibility; access to financialinfrastructure; leapfrogging the digital divide; multi-functionality;

scalability; and minimising fiduciary risk and fraudulent access.60

Acknowledging the considerable benefits that can be derivedfrom integrating ICTs into the delivery of social protection, theuse of information and communication technologies neverthelesspose a number of risks to beneficiaries right to privacy, asextensive and sensitive information is collected, analysed anddisseminated about them. In particular, the use of electronicManagement Information Systems (MISs) to collate and generate

information about social protection beneficiaries and informtargeting, management, reporting and analysis raise seriousconcerns. MISs facilitate the gathering and storing of extensiveamounts of personal data in what are often insecure or high-riskenvironments. Where donors or development agencies administerthe scheme, and where private-public partnerships are integratedinto the scheme, the potential for abuse of beneficiaries personalinformation is high. There is some confusion around the ownershipand use of sensitive personal information collected by socialprotection programmes; these concerns are particularly serious

in low-income countries where data protection laws are weak,or non-existent.

Similar concerns exist with the move away from cash or in kindtransfers and towards electronic transfers by aid agencies. The card or mobile enabled conversion of cash into electronic moneyhas been a hugely successful advancement in the provision ofsocial protection transfers in developing countries. However,numerous risks arise due to the sharing and transfer of personal

information with third parties.


20/77

20/80


ACF International

2,540,878(25%)

IRFC/National Societies

2,222,303(22%)

OXFAM

1,228,872(12%)

World Food Program (WFP)

1,043,174(10%)

Concern Worldwide

923,311(09%)

Others

2,303,512(22%)

Projects

330Beneficiaries

10,262,050

Cash transfer programmesmapped by the Cash

Learning Partnership (CALP)

Section 3


21/77

Number of beneficiaries per organisations:

22%25%

22%

12%

10%

09%

Source: CALP Cash Atlas.

Section 3 Aiding Surveillance Privacy International


22/77

22/80


The potential

Increasingly, social protection programmes in developing

countries are making the transition from paper to electronicsystems. Complete transition to fully-integrated electronicsystems remains elusive in many low-income countries,particularly those hampered by difficulties in access toelectricity, internet and mobile phone networks. Generally,however, donors and development agencies are encouragingthe adoption of electronic systems in social protection pilotsand supporting the migration of social protection informationto centralised single registries.

Within the literature on social protection, MISs are identified as anintegral part of the administration of social protection programmes,enabling the collation and application of information related tothe various components of the scheme, including those related toregistration, conditions, targeting, payments, grievance systems,and graduation. MISs collect and collate an extensive amount ofdata, particularly in those social protection systems which requirecompliance with programme conditionalities (which often relate toattendance at health or education services)as a prerequisite for

receipt of benefits.61The following table is illustrative of the typesand amount of personal information collected about a socialprotection beneficiary.

HelpAge International notes that additional information on recipientsdegrades in accuracy as soon as it collected, as people leave thehousehold, children are born and assets are sold or purchased.Much of the information, in addition to being inaccurate, is of littleuse, and is extraneous to the determination about beneficiary

eligibility, according to HelpAge.

Data to populate the MISs is collected through a variety offorms, mostly electronic (using laptops and mobile devices).62Once digitised, MISs allow for increased flows of data to otherinstitutions. Research shows that the adoption of MISs toadminister social protection information increases the ability of

3.2

Section 3


23/77

23/80


programmes to send data directly from communities or districtsto databases held in the capital cities.

Linked to the adoption of MISs is the move towards a SingleRegistry of social protection programmes in each country. Thedrive towards a Single Registry is inspired by Brazils CadastroUnico which aims to build a database of the entire poor populationof Brazil; it now holds data on the declared incomes of 16 millionhouseholds and uses an unverified means test for targeting.The main user of the Cadastro Unico is the Bolsa Familia scheme,but it has also been adopted by nine other schemes.63

The availability and persistence of this information means that ifeffective, it could provide a single source of information on largepopulations, available to numerous stakeholders with differentiatedlevels of access, as has been the case in Chile.64It is therefore opento re-use for other purposes and by other State and potentiallynon-State entities. In Kenya, for example, the government is rollingout an Integrated Financial MIS that integrates social protectionpayments with all other electronic payments made through theCentral Bank electronic payment system.65

Electronic transfers also have huge beneficial implications forhumanitarian assistance. By providing a secure and simple methodof potentially life-saving transfers to vulnerable groups, electronictransfers could make social protection programmes more efficientand effective, while at the same time creating a new resource ofinformation about how money is spent. Organisations such asConcern Worldwide have been quick to take up electronictransfers, working with Safaricoms M-PESA system, and theCash Learning Partnership, a consortium of non-governmental

organisations such as Oxfam and the Norwegian RefugeeCouncil, are currently investing considerable research into pilotingguidelines and a Code of Conduct related to the use of e-transfers.

Section 3


24/77

Section 3

PRIMARY MONITORING INFORMATION

ADDITIONAL INFORMATION ONAPPLICANT/RECIPIENT

Marital status

Educational attainment

Additional address details(e.g. family name, name known by)

INFORMATION ON HOUSEHOLD MEMBERS

Number of members

Date of birth

Sex

ID number

Marital status

Single/double orphan

Relationship to beneficiaryand/or household head

Educational attainment

INFORMATION ON DWELLING AND ASSETS

Water source

Sanitation

Landholding size

Land tenure

ACCESSIBILITY OF SERVICES

Distance to health clinic

Distance to primary school

Distance to secondary school

Distance to pay-point

SECONDARY MONITORING INFORMATION

Occupational status

Disability

Disability status

Occupational status

Grade enrolled at school

Health status

Description of dwelling

Type and number of animals

Car

Bicycle

Agricultural implements

Etc.

Source: HelpAge International, 2011

Aiding Surveillance


25/77

25/80


The problems

The collation of extensive and sensitive personal information

in an MIS lends itself to a number of challenges in respect ofprivacy and data protection. These include:

Accuracy of data: Multiple obstacles exist to collecting accurateand comprehensive data in situations in which social protectionprogrammes are administered, including the geographicalremoteness of target communities, social exclusion anddiscrimination, lack of literacy, and the absence of formalregistration records. By enabling the digitisation and indefinite

preservation of potentially inaccurate data, MISs risk reinforcingand institutionalising such inaccuracies, which may be impossiblefor beneficiaries to correct.

Security of data: Ongoing technical support and maintenanceof a system is key to ensuring security, and yet requires a level ofexpertise and capacity that may not be present in donor-runor pilot social protection schemes. Systems that involve thetransfer of data via telecommunications networks face additionalthreats in the absence of encryption or where State authorities

are conducting communications surveillance.

Misuse of data: Any personal information contained in MISs isvulnerable to fraud or theft, as well as transfer to third parties.The higher the sensitivity of the data data, for example, thatreveals or could be paired with other data to reveal ethnicity,religion or political affiliation the more vulnerable it is.

A further challenge of adapting MISs to social protection

programmes is ensuring that the technologies deployed areappropriate to the relevant culture and context. A study of thedevelopment of an MIS for the distribution of social protectionbenefits in St. Kitts revealed that it is necessary to understandthe contexts in which data is collected and used to ensure thatthe [MIS]will fit within the users work environment and beuseful to them.66Simple assumptions inherent in the design of

3.3


26/77

26/80


technology such as the requirement to enter addresses that followa predefined format may undermine the utility and effectivenessof MISs in developing countries. It is estimated that in developed

countries approximately 25 per cent of MIS projects are failures,and up to 60 per cent have significant undesirable outcomes;in developing countries, this number is likely to be significantlyhigher.67A failure to take into account cultural contexts may bea contributing to such failures, which also stem from factorssuch as cost overruns, insufficiently trained staff, and inadequateprocesses. Early studies in this field showed that almost all WorldBank-funded MIS projects in Africa were reported as partial failure.68

In order to ensure that an MIS takes into account the particularcontext of the country, it will most likely require a custom-madesolution. However, most social protection programmes haveneither the resources nor capacity to do so cost-effectively,and as such rely on generic MIS solutions. Kenyas Urban FoodSubsidy relies on Microsoft Access, for example, while Mauritiususes Oracle, South Africa uses Adabas and the Hunger Safety NetProgramme and OVC-CT in Kenya use Microsofts SQL Serverdatabase.69Research shows that a generic approach has seriousdrawbacks and is unlikely to be successful.70

When social protection programmes use generic MISs this raisesadditional questions about who might ultimately have access tothe data. The role of private corporate entities in social protectionprogrammes more broadly is also an issue. The situation inSwaziland is apposite the government is in negotiations aboutcontracting Standard Bank and the SwaziPost to administer thecountrys Old Age Grant. Should the scheme come to fruition,Standard Bank will hold a separate database with information on

all 45,000 beneficiaries.71

Similarly, the Dowa Emergency CashTransfers project in Malawi was administered by ConcernWorldwide contracting the Opportunity International Bank Malawiand the Malawi Police Service.72Issues around informationgovernance will arise: who owns the information, who is responsiblefor problems and who is mitigating any risks of abuse?


27/77

27/80


The administration of electronic transfers lies at the heart ofthis challenge. E-transfers rely on the private sector to providethe telecommunications and financial infrastructure, and to

design and maintain the banking and mobile systems upon whiche-transfers rely. Electronic cash transfer systems are often runby small NGOs on a pilot basis without concrete structures,extensive legal expertise or sufficient resources to ensure thatthird party contracts are rigorously analysed and complied with.The likelihood that beneficiary data is being shared and analysedby third parties is thus increased.

The beneficiary data collected for e-transfer programmes is often

more extensive than that gathered in conventional aid distributionsand is necessarily shared with commercial partners who assist in thedistribution of cash via new technological means. The developmentof sophisticated databases, the sharing of those databaseswith third parties, and the lack of technical and operationalsecurity around the collection, use and sharing of data all createa heightened risk framework, at the heart of which are the verypeople agencies seek to support.

The risks of deploying MISs and e-transfers in social protection

programmes are heightened by the absence of legal frameworksand safeguards to regulate the use of data collected under theauspices of such programmes. In most developing countries,data protection legislation is weak or non-existent. Manysocial protection programmes are established ad hoc, as pilotprogrammes by development and humanitarian agencies, or underthe ambit of bilateral aid agreements, without accompanyinglegislative or regulatory frameworks. This means that the rightsof the beneficiaries in the programme are unprotected, and the

administrators of the programme have wide discretion whendealing with beneficiaries personal information. In any event,given that many programmes are the result of a collaborativeeffort by multiple stakeholders including donors, governmentactors and international NGOs there are serious questionsabout accountability, transparency and avenues for recoursefor beneficiaries.


28/77

28/80


Digital identity registrationand biometrics

The promise

Ensuring that development and humanitarian aid reachesthose for whom it is intended is a perennial challenge forforeign aid donors and international funding organisations.Obstacles to delivering aid include not only securityrisks73and lack of infrastructure (airports, roads and otherfacilities),74but also the difficulty of identifying and targeting

intended beneficiaries.

Increased pressure to focus aid where it is needed most, and tomonitor aid programmes has resulted in a strong push for greaterinformation on recipients. The benefits to development policy oftargeted approaches are clear: properly identify the individualsand the groups that need assistance, and programmes will becomemore effective and efficient. As the US Government AccountabilityOffice (GAO)framed it in a September 2012 report on targeting of

food aid, effective targeting is important to maximize the impactof limited resources, with a particular emphasis on the quality ofdata used to identify and reach recipients.75

A significant challenge in targeting is ensuring that there aresufficient amounts of information on the target populations toensure that the determinations made are necessary, proportionate,and critically assessed, and that aid delivery can be tracked andmonitored to assess its effectiveness. In order to begin to addressthese issues, some donors have begun to use technology

to support identification and registration. Development andhumanitarian initiatives related to providing refugee assistance,delivering social protection or food subsidies, and improvingdemocratic institutions, particularly electoral reform, have allbegun to integrate digital identity registration.

4.1


29/77

29/80


Case study: UNHCR biometric identity registration in Dijbouti,Ethiopia, Kenya and Malaysia

UNHCR biometric identity registration in Djibouti, Ethiopia,Kenya and Malaysia

The UN Refugee Agency (UNHCR)has long used databasesto collect and manage information on refugees, and hasissued refugees with a form of certification of their status.In recent years, UNHCR has begun to deploy biometricidentification systems to register refugees, and check theiridentity and status for aid disbursement. Pilot schemes were

initiated in Eastern Africa and Asia in the mid-2000s, and inOctober 2012 UNHCR announced that it was to begin usingbiometrics in Senegal and South Sudan.

A field study conducted by the authors in 2008 witnessedthe considerable problems being experienced with respectto UNHCRs deployment of a biometric system. The primaryconcern was the systems reliability: UNHCR had procureda fingerprinting system that was not designed for largepopulations, and particularly not for large populations that

did not have well-defined fingerprints.

UNHCR staff members were unaware of this problem andlacked guidance on how to use the system in the field: variousfield operations were using the system differently, somefingerprinting adults of all ages, young people, and evenbabies, presuming that the system would work. The systemwas erratic; it worked sometimes on someone, and sometimeson that same person it would not work even moments

later. But for UNHCR, it was a perceived success: staff hadhigh confidence in the system, and it was a useful tool forcommunicating with host governments that UNHCR wastaking fraud seriously.


30/77

30/80


The recording of identity into registers is not new; identityregistries exist in many forms in many countries. Indeed, themaintenance of an effective system of identification is arguably

essential for the development of individuals legal identity, to thedistribution of social services,76and to the realisation of the rightto identity registration at birth, enshrined in the United NationsConvention on the Rights of the Child.77However technologies arechanging the impact and importance of identity registration in twoways. First, they are enabling the digitisation and centralisationof these registries, their use across government services,and the continual checking of identity. Second, technologicaladvancements have facilitated the capture, processing and

retention of biometrics, physical traits of individuals includingfingerprints, facial scans, iris scans, or even DNA. These relativelyunique characteristics can provide identifiers across systems,and even across borders, tracking individuals across contexts,allowing for the reuse of information. They also make sharing,linking and cross-checking information faster.

Proponents of digital identity registration and the tying of identityto biometric information maintain that such systems can help toempower individuals by giving them legal identity and connecting

them to services. Biometric identification, it is argued, is moreaccurate and thus its employment more likely to forestall identityfraud and improve the transparency and accuracy of electoralprocesses and access to public services. According to thedirector of Indias universal biometric identification scheme, sucha scheme can be transformational, and solve the most basic ofdevelopmental challenges.78By offering a solution to the absenceof traditions of birth registration and accompanying infrastructure,biometric identification systems provide for the opportunity

to uniquely register a large population of people, and in turn,administer entitlements. By connecting data to a unique personalidentifier such as a fingerprint or iris scan, biometrics avoid theopportunities for forgery associated with other formsof identification.


31/77

31/80


The potential

Biometric identification systems are used to record and

identify social protection beneficiaries in at least 15 cash-transfer programmes: Pakistan, Afghanistan, the DemocraticRepublic of the Congo, Malawi, South Africa, India, Ghana,Namibia, Botswana, Kenya, Nigeria, Iraq, Philippines, Bolivia,and Indonesia.79Such systems tie the biometric informationof the beneficiary to the information held about them bythe social protection programme. Benefits are disbursed viashopkeeper-operatedpoint-of-sale devices, which verifythe fingerprint scan, connect with the central database, and

transfer funds into the shop account that are immediatelypassed along in cash to the recipient.

Biometric technologies are particularly prevalent in Africa, andare spreading; estimates put biometrics technology in at least34 countries in Africa. This primarily takes the form of biometricnational identity cards or biometric voter registration systems thatincorporate Automatic Fingerprint Recognition Systems (AFIS),fingerprints being the dominant form of biometric data collected.

Prominent instances of biometric identification systems include:

Democratic Republic of the Congo: Biometric informationis a key element of the Disarmament, Demobilisation andReintegration Programme (PNDDR)in the DRC, establishedin 2004 and co-funded by the World Bank. The programmedisburses 13 cash payments over the course of a year toex-combatants. Biometrics in the form of iris scans, asfingerprints were unreliable for ex-combatants with calluses

on trigger fingers were introduced in 2006 to enrol 110,000individuals.80Beneficiaries visit one of ten mobile paymentteams in rural areas, have their irises scanned and receivetheir payments.

India: The state of Andhra Pradesh was one of the first touse biometrics to deliver government payments, partneringwith FINO, an Indian technology company which designed a

4.2


32/77

32/80


platform based on biometric identification to link rural citizenswith the formal banking system.81

Pakistan: The Watan card an identification card containingbiometric data that can be credited with social protectiontransfers was introduced after the 2010 floods by the NationalDatabase and Registration Authority, and used as a meansof transferring National Flood Relief Grants to over 1.5 millionvictims, in a programme jointly administered by the governmentand UNHCR.82

South Africa: One of the oldest systems of biometric registration

in the world exists in South Africa; the government begancollecting the fingerprints of non-white citizens in 1925 for thepurpose of racial registration. In 1992, the province of Kwa-ZuluNatal worked with Net1, a South African company, to set upbiometric technology to enable the payment of social protectiongrants to pensioners. The system continues to be extended,and now distributes grants to over 15 millionbeneficiaries.In 20122013, a new system provided by Net1/Cash PaymasterSystems captured the biometric information of more than20 million South Africans as part of a new national social

protection payment system which was aimed at reducingfraud and corruption.

In many cases, the technology is procured from foreign companies,many of them European. Unlike in other development sectors wherethere is now a drive towards local sourcing, biometric programmesoften involve spending money designated for developing countrieson Western high-tech firms.

Development agencies and bilateral donors have played alarge role in supporting biometric initiatives. In 2011 the UNDevelopment Programme (UNDP)provided 26 per cent ofits funding towards fostering democratic governance in thedeveloping world.83In Africa alone, through the United NationsDemocracy Fund, the UNDP has funded biometric voterregistration in Benin,84Cape Verde,85the Comoros Islands,86

Section 4


33/77

33/80


Democratic Republic of Congo,87Sierra Leone,88Togo,89andZambia.90Other examples of development funding for biometricsystems include World Bank funding for registration of the urban

poor in Benin91and Kenya.92USAID has funded biometric systemsin Malawi 93and Guinea,94and played a large role in supporting theregistration of 14.3 million voters using biometric voter registrationtechnology in the lead up to the 2013 Kenyan elections.95

The costs of deploying and operating these systems are significant.In Mozambique the cost of the national identity cards, contractedto Face Technologies, was US$15 million.96UNDP funding tobiometric registration and machines in Sierra Leone was US$18

million for the 2012 elections.97

The contract between Ugandaand Muhlbauer group was 64 million.98In Ghana the costs wereestimated at US$100 million.99

The problems

Biometrics, whether based on face, finger, iris, DNA, or someother physical or genetic characteristic, are in many ways just

another form of personal information, and their registrationand connection with identification thus give rise to questionsof privacy and data protection.

Yet we cannot ignore the ethical dimensions. Identificationregistration systems have problematic legacies. In Rwanda, thecolonial racialisation of the identities of Hutu and Tutsi contributedto the increasing polarisation of the two groups in the postcolonialperiod, leading to the 1994 genocide.100The use of identificationcards was a key administrative component of this as they allowed

differential access to the two groups entitling Tutsis to far moreextensive political and social freedoms than Hutus.101The Belgianscolonial approach was to institute an ethnic classification,involving such modern scientific methods as a measurement ofnose and skull sizes, and required this information on mandatoryidentity papers.102

4.3


34/77

34/80


Since 2007 Rwandas National Identification Department hascreated a permanent civil and voter registry, and citizens datais held in a central and permanent database.103While there is no

ethnicity information on the new cards, they do contain biometricdata the fingerprints of approximately 9.2 million Rwandancitizens have been collected and stored.104Although the useof biometric registration has since been greeted positively inRwanda,105the serious nature of the problem of political abuseof biometrics becomes apparent in this context.

The artificially constructed identities of Tutsi and Hutu were usedto secure political, social and economic benefits. It is possible to

imagine categories of identities relating to fingerprints beingsimilarly constructed and being used to the advantage of politicalor criminal groupings.

The use of biometrics in South Africa also raises questionsconcerning the human dimensions of the use of biometricidentification systems. Although a key mechanism for thefunctioning of citizenship in the country, the national populationregister was also the administrative and ideological cornerstoneof apartheid. The 1950 Population Registration Act required

peoples identity numbers to refer to ethnicity. Although ethnicityis no longer incorporated as part of identity documentationin South Africa, this history raises important questions aboutidentification systems that have the potential to be used fordiscriminatory purposes and social sorting.

Few registration systems are now considering including ethnicityinformation because of these lessons. But the inclusion ofbiometrics and additional biographic information raises new

concerns. The linkability of biometrics increases the likelihoodof their expansion and re-purposing in other environments (inthe criminal justice or immigration systems, for example)or forother purposes unimagined at the time of their collection. Oneof the predominant reasons why digital identification systems,particularly those containing biometrics, have faced resistancein developed countries is the potential for scope creep: once

Section 4


35/77

35/80


Case study: Aadhaar Unique ID project (UID) in India

Recent experiences with the UID project in India demonstratethe complications that can be faced in deploying biometricidentification systems. In 2009, the UID Authority of Indiawas established to carry out the UID scheme with theobjective of issuing every resident in India with a uniqueidentification number based on their biometrics, designedto eliminate duplicate identities and authenticate individualsin a cost-effective way. Implementation of the project hasbeen conducted since 2010 in the absence of legislation.

The UID was initially designed to be an identification tool toauthenticate and provide services, adoptable by any platformin a consolidated manner. But without clear limitations on itsuse, the number has been adopted by various services andplatforms for their unique purposes including identification,linking, and tracking individuals in various systems. Inthis way the UID number has expanded from being just anauthenticator, to being an identity and a tool for servicedelivery, and increasingly mandatory for access to many

services. For example: the Indian government has required that citizens have

a UID number to purchase cooking gas, issue an open-government request for information, and register vehicles.

the High Court has directed all police stations inMaharashtra to record the UIDs of accused individualsand witnesses filing an incident report.

railways are proposing to use the UID database forbookings and validation of passengers;

the city of New Delhi is implementing a scheme calledSaral Money that allows individuals to open bank accountsonce they have stated their UID number.


36/77

36/80


the Rajasthan Government has made it mandatory foremployees to have a UID number and has linked thenumber to employee salaries. Yet, infrastructure issues

including a lack of available machines has preventedindividuals from enrolling for the number.

The system faces numerous serious challenges, including:

many rural workers, elderly, and poor individuals do nothave readable fingerprints. It has been reported thatoften agencies are simply refusing to enrol such individuals,thereby excluding them from the service and all the

subsequent uses (and benefits)of the UID. enrolment centres are overcrowded without proper facilities.

duplicate numbers have been issued and some enrolmentagencies have been blacklisted for fraudulent practices.

Malu. B. The Aadhaar Card What are the real intentions ofthe UPA Government? DNA. February 18th 2013. Available at:http://www.dnaindia.com/blogs/post_the-aadhar-card-what-are-the-real-intentions-of-the-upa-_1801080-all. Last accessed:

February 28th 2013.

Misra. U. Inside the Direct Cash Transfer Debate. Forbes India.January 2013. Available at: http://forbesindia.com/article/briefing/inside-the-direct-cash-transfer-debate/34510/1

Plumber, M. Make UID numbers must in FIRs: Bombay HC. DNA.October 2011. Available at: http://www.dnaindia.com/mumbai/report_make-uid-numbers-must-in-firs-bombay-hc_1603127. Lastaccessed: February 28th 2013.

Times of India. Govt tries to ramp up Aadhaar enrollments,

but centres ill-equipped. March 3rd 2013. Available at: http://articles.timesofindia.indiatimes.com/2013-03-03/jaipur/37409682_1_uid-registration-uid-card-aadhaar

UIDAI Strategy Overview. Creating a Unique Identity Number forEvery Resident in India. April 2012. Pg.2. Available at: http://uidai.gov.in/UID_PDF/Front_Page_Articles/Documents/Strategy_Overveiw-001.pdf

Section 4


37/77

37/80


collected, biometrics can be re-used for a variety of otherpurposes.Therefore, a system that is designed for the purposeof disbursing aid and entitlement services will soon be used for

verifying citizenship and age, and biometrics may be checked andcompared with those for policing purposes.

From a privacy perspective, some biometric applications are moresensitive than others: for example, photographs enabling facialrecognition and DNA records facilitating genetic profiling canassist the creation of racial and ethnic profiles. The invasivenessof collection also has a bearing on the privacy impact of thetechnology. DNA requires intimate contact with the individual,

and even submitting to facial recognition technology may requirethe removal of clothing. Studies have indicated that there is someconcern from users about the requirement to physically toucha fingerprint scanner, or even cast ones eyes into a biometricscanner for retina or iris recognition. In some circumstances, facialimages can be collected without the consent or knowledge ofthe individual; fingerprints and DNA can be collected from latentprints or samples left behind on objects and linked with otherdatabases and activities.

The predominant form of biometric recognition used in developingcountries is fingerprinting. However, fingers are vulnerable andprints are not always easy to read.106For example, fingerprintscanners tend to fail more frequently on women in developingcountries, as their fingerprints have been degraded due to manuallabour. Some fingerprint recognition systems may also havedifficulty in registering the fingerprints of the elderly, those withsmall or fine fingerprints, and fingerprints that may have worndown such as those of manual and rural workers. This can result

in high failure to enrol rates meaning that a number of individualscannot be read by the technology and therefore cannotparticipate in the registration that is taking place.

Digital identification registration systems increase the likelihoodthat processes become data dependent, and in turn, thatdeterminations are driven by such data. The adoption of biometric


38/77

38/80


technologies means that sensitive personal information on entirepopulations can be collected and processed rapidly, anddecisions can be made with reference to digital profiles and

aggregated data, the integrity and veracity of which is difficultto establish or safeguard. Information and data is not value-free,and discriminatory judgments can become accepted andinstitutionalised through the use of automated systems.Individuals quickly become reduced to a set of knowable andmeasurable facts that may not necessarily represent them or theircircumstances accurately. With the advent and proliferation ofthe EURODAC biometric database system for identifying asylum-seekers and irregular migrants, submitting to biometric registration

has become a de facto pre-requisite to claiming asylum. Asylumseekers and refugees are reduced to someone with a file, whosebiometrics need to be verified in order to gain access to, or beprevented from wrongly accessing services. When the biometricssystems do not accurately function, the refugees status is therebycalled into question sooner than the technology. This may lead tothe further marginalisation of vulnerable individuals, other humanrights violations, and exclusion from vital aid. In September 2013,for example, 6,500 refugees in the Mbera camp in Mauritaniawere denied access to refugee assistance because of problems

with the biometric registration system.107

Case study: Biometric voter registration in the DRC

The DRCs first democratic elections in four decades wereheld on 30 July 2006. In support of the 2006 election processthe international community donated US$460 million to the DRC.La Commission lectorale Indpendante (CEI)of the DRCdecided to biometrically register voters for the elections.The UNDP oversaw the procurement process and two contractswere awarded to European companies to institute biometricregistration. Zetes was awarded the contract for 10,000biometric registration kits, at a cost of US$40.16 million andSagem, a French company, was charged with removing


39/77

39/80


duplicates in the system. An additional US$58 million wasspent on the operation costs.

The biometric system was implemented in a context in whichthere was:

no reliable electoral list or any demographic data from1984 onwards.

a lack of basic infrastructure; in the DRC there are only42,000 fixed phone lines for a population of over 73 million,and only 9 per cent of the country has access to electricity.

no centralised fingerprint matching system within thesystem itself, meaning that checking for duplicates withinthe registration system could not be carried out withinthe DRC itself, but instead by the European company thatdesigned the system.

a high degree of machine malfunction, and systematicflaws in the system that required its redesign.

Despite these problems, biometric registration went aheadand was used in the 2006, 2009 and 2011 elections. Yet thecountry has remained plagued by undemocratic institutionsand claims of electoral fraud; in 2011 violence broke out whendoctored ballot papers were found. A leaked report fromZetes recorded that there had been more than 700,000 doubleregistrations on the biometric system.

EISA, (2010), Voter Registration in Africa: A ComparativeAnalysis, Case study on DRC by Akumiah, H., p.57-102, available

at: http://www.eisa.org.za/PDF/vrafrica.pdf

McElroy, D., (2011), UK pays 22.5 million for questionableDemocratic Republic of Congo election from The Telegraph[online] 16 October 2011, available at: http://www.telegraph.co.uk/news/worldnews/africaandindianocean/democraticrepublicofcongo/8830144/UK-pays-22.5-million-for-questionable-Democratic-Republic-of-Congo-election.html


40/77

40/80


While biometric identification systems may offer significantopportunities for development, equally they may not be suitablefor countries where there is very little communications or transport

infrastructure. Biometric identification systems require a constantelectricity supply, and registration kits include a computer andprinter at a minimum. They may require reliable transportation, highlytrained personnel to operate the systems, a network connection, anaccessible and accurate civil or voter registry, webcams, extensivedata storage facilities or any number of additional componentsin order to operate effectively. Just as with those developmentprojects that provide infrastructure to a community without thetools, expertise and capacity to maintain and integrate such

infrastructure, so too are biometric projects which do not takeinto account the local context doomed to fail.

Although many biometric identification systems have beenadapted for use in difficult conditions (primarily by encasing thekits in hardy coverings, and budgeting for back-up electricitygenerators)the nature of this technology means that it can befragile and susceptible to damage, or attractive to thieves. Anyinvestment into a biometric identification system simultaneouslyrequires investment into the infrastructure required to support

and protect these systems. However, research suggests thatdespite claims by proponents, biometric systems are not infallible,and systems and processes around biometric registration aresusceptible to fraud, forgery and corruption. Research intomedical record registries, for example, reveals that leaked orstolen medical information has not only been sold for profit, buthas been used by government agencies and to publicly shamepolitical figures.108

The possibility for misuse of biometric identification information ishigh, and the potential for harm to follow is very real. The ability fordigital identification systems to be used as a means of surveillancehas been recognised by producers of biometric technologiesand even emphasised as a selling point to make the technologyattractive to repressive regimes. The director of one firm involvedwith the deployment of Egypts ID system under Mubarak, and


41/77

41/80


funded by the Danish Aid Agency DANIDA, recognised that [thetechnology] could be used for surveillance We can easily designa program for the ID card which enables surveillance of users

internet activities or conversations on Skype This is business,we sell to those who are interested. If I was approached by Iran Iwould sell to them.109

Biometrics is a growth market for technology companies,particularly in developing countries. Zetes, the supplier ofbiometric technology to Cape Verde, Ivory Coast, DRC and Togostates on its corporate website that the interest of governmentsand international institutions in biometrics is growing. They note

in the Western world, the use of biometrics has been raising someprivacy concerns. That doesnt seem to be the case on the Africancontinent, where biometrics are regularly used.110

As the role of the private sector in providing biometrics systemsto the governments of developing countries continues to expand,the problematic nature of such relationships becomes increasinglyclear. Procurement contracts have been questioned in the case ofthe Muhlbauer group in Uganda,111Semlex in Mozambique,112Net1in South Africa,113and Giesiecke & Devrient in Cameroon.114In

Mozambique, several stakeholders accused the National ElectoralCommission of a lack of transparency.115Much of the equipmentprovided by Mhlbauer to Uganda has been lost or broken, andonly 400 ID cards had been produced since the contract beganin 2010 up until late 2011.116Recently, Mastercard and Nigeriaannounced a shared initiative to deploy a shared national ID thatwould combine biometric functionality with electronic payments.There is little information on how information will be managedbetween the company and the government.117


42/77

42/80


Mobile phonesand data

The promise

The arrival of mobile telephony in developing countries hasplayed a crucial role in the success of many developmentinterventions over the past ten years. Mobile phones havenot only greatly improved opportunities for communicationand expression, they have enabled financial empowerment,provided access to information and services, and revolutionised

the collection and recording of information in humanitariandisasters. Systems such as Kenyas M-PESA mobile moneysystem, which allows individuals to bypass traditionalfinancial infrastructure and access and transfer money by SMS,has greatly reduced the financial exclusion of vulnerablegroups, improving their ability to save money and accumulateassets. In the first three months of M-PESAs operation, 11,000people registered for the service, and nearly US $6million wastransferred; today it is used by a quarter of the population,

some of whom had not previously used mobile phones orowned bank accounts.118

Linked to the proliferation of mobile phones in developingcountries are initiatives designed to use the data generated orcollected by mobile phones to conduct analysis about trends andevents that might inform future development and humanitarianinitiatives. Big data the amassing and analysis of high volumesof digitial data to uncover new correlations is taking thedevelopment world by storm, facilitated by the rapid reproduction

of the quantity and diversity of data generated by digital activitiesconducted on mobile phones, particularly smart phones calllogs, mobile-banking transactions, online user-generated content,online searches, satellite images, etc. Algorithms are applied todevelop intelligence on people, groups, and events and places.With enough data, the theory goes, we even can try to predictbehaviour based on past activities.

5.1


43/77

43/80


International agencies and organisations such as UN Global Pulse,119UN Economic Commission for Latin America and the Carribbean,120OECD121and the World Economic Forum122have all sung the

praises of data as a tool to accelerate development, reducepoverty, spur innovation, and improve accountability andtransparency. A recent report of the UN High Level Panel on thePost-2015 Development Agenda went so far as to call for a NewData Revolution, drawing on existing and new sources of data

to fully integrate statistics into decision making, promote openaccess to, and use of, data and ensure increased support forstatistical systems.123


44/77

44/80


Global mobile data trafficgrowth & forecast

(terabytes per month)

North America

Central &Eastern Europe

Latin America

Western Europe

Middle East& Africa

Asia Pacific

Source: Cisco, Cisco Visual Networking Index: Global MobileData Traffic Forecast Update, 20122017, February 2013; CiscoVisual Networking Index: Global Mobile Data Traffic ForecastUpdate, 20092014, February 2010.

2,000,000

4,000,000

6,000,000

8,000,000

10,000,000

12,000,000

0

2012 2013 2014 2015 2016 2017


45/77

45/80


The potential

The use of mobile phones in developing countries to collect

or generate data, and the subsequent analysis of such data,has the potential to assist in development and humanitarianinitiatives in multiple ways:

Health services: mobile phones are used as a means todispense health information and connect individuals to healthservices. mHealth for Development, founded by the UNFoundation and Vodafone, supports the use of mobile phonesto send SMS text alerts to enable patients to adhere to their

prescriptions, and to train health care workers. In Ghana, theMillennium Villages Project provides diagnosis and treatmentsupport to rural health workers.124

Health trends: big data analysis of mobile phone location andsocial media trends is used to track public health trends. Suchtools have been employed with success in Haiti: research fromthe cholera outbreak there identified Twitter as a useful sourceof information about the extent of the outbreak. By analysingover 188,000 tweets spanning a three-month period, researchers

were able to monitor the outbreak and its progress much fasterthan through government processes that involved surveyinghospitals and clinics. Other researchers conducted analysis ofcell tower data to plot the location of populations fleeing fromthe outbreak.125

Crisis mapping: by using data submitted crowd-sourced frommobile phone users on security or humanitarian crises, crisismapping platforms are able to map incidences of violence or

disasters. A prominent example of such a platform is Ushahidiin Kenya, funded by, among others, the Ford Foundation andMacArthur. Ushahidi played an important role in the post-earthquake response in Haiti, enabling the creation of a crisismap of urgent humanitarian needs.126

5.2


46/77

46/80


Infrastructure and services: big data can be used to mapinfrastructure and the use of public services like transport.Telecommunications company Orange recently opened a big

dataset of 2.5 billion anonymised text messages and phonecalls from Cte dIvoire, enabling researchers to analyses andredesign bus routes in the country.

Reporting: mobile governance projects such as MexicosCitivox and Indias Kerala State IT Mission enable citizensto register to vote or report crime and corruption via theirmobile phone.

Conflict prevention: emerging research argues that big data canbe used to prevent conflicts, by distinguishing digital patternsand interpreting them in the applicable socioeconomic andpolitical context, or studying cause and expressions of concernsand stress in a given community.127

Monitoring and evaluation: big data can be used to analyse largepopulations and report back to funders on the effectivenessof programmes.

The problems128

In an age of widespread communications surveillance byboth State and non-State actors, using mobile networksto transmit sensitive data is inherently risky. Developmentand humanitarian initiatives that use mobile phones tocollect or generate information thus risk such informationbeing exposed to potentially malevolent third parties; or

fraudulently amended or misappropriated.

Mobile health is an area around which particularly serious concernsarise. A recent report by TrustLaw, in collaboration with the mHealthAlliance, recognised that the lack of comprehensive data protectionand privacy protections in developing countries has impeded theeffective expansion of mHealth initiatives.129Numerous practical

5.3

Section 5


47/77

47/80


barriers stand in the way of mobile health initiatives; althoughmobile phones are arguably the success story in the domain ofinformation technology and development, their diffusion still is

not universal. Not everyone has a mobile phone. Often phonesare shared by families; in some contexts, the dominant male in thehousehold (usually the father)owns the phone. In this scenario,the use of mobile phones for notifying individuals about, forexample, a test result, to report incidents of domestic violence,or to provide reminders about an appointment of which theirfamily members were not previously aware is a complicated affair.What sort of information should be disclosed in the text messageitself? While it may be possible to exclude specifics about a

disease or medication, in certain areas the mere fact that one isbeing contacted by a health actor can be stigmatizing. Therefore,some eHealth systems have started obfuscating these messages,using codes such as sport scores or messages from friends tocommunicate sensitive health data.

However, there are other complications to the use of mobilephones for health. Across the globe, governments are requiringcitizens to register their SIM cards with personal information.An example of this is the case of VidaNet, a HIV patient reminder

system in operation in Mexico City, which is currently strugglingto provide a privacy-friendly service as the country enforces anational SIM registration program.

Not only disseminating information is problematic; gathering andanalysing big data sets of mobile phone activity also presents aserious challenge to the protection of individuals. Digitising dataand pairing it with multiple other data sources can result in themosaic effect, allowing for data elements that in isolation appear

non-personal or innocuous to be combined to enable the detailedprofiling of individuals. It imagines that personal information is aresource that can be mined and disclosed by the organisationwithout any consideration of the wishes of the individual.

Proponents of big and open data argue that their information isanonymised, and the analyses are about the aggregate, not theindividual. The serious problems with data anonymisation130and


48/77

48/80


the potential for de-anonymisation have been well publicised andcontinue to plague the big and open data movements, despiteassurances by regulators that such risks can be mitigated.

The problems of anonymisation are enhanced by the lack ofsafeguards and standards inherent in data for developmentinitiatives. International consensus on detailed data protectionstandards remains a work in progress, data protection legislation isstill largely absent on the African continent, and few developmentand humanitarian organisations have self-standing data protectionand privacy policies to guide their work in developing countries.As the UN itself admits, while private-sector organisations

and [g]overnment regulators have been grappling with thisissue for almost a decade, humanitarian organisations appearfurther behind.131In the absence of strong legal safeguards andaccountability institutions individuals in developing countries havelittle recourse against the violation of their privacy.Data is not context free. Developing countries are also plagued byhistorical divisions, ethnic conflicts and other social and culturalvulnerabilities that heighten the risk that big and open data willbe misused. Discrimination or persecution could easily be the

result of de-anonymisation of big data pertaining to, for example,electoral trends, public health issues, political activity or location.Call and text message records held by the private sector, forexample, were used by the Egyptian authorities to track downand convict protesters in the aftermath of anti-government foodprotests in 2008.132The risk of the misuse of personal datais heightened when data is open and thus accessible by anyone for any reason. Even the open digitisation and publicationof seemingly banal information can have adverse effects in

Pakistan, for example, the publication of locations of fooddistribution points and clinics led to threats to aid workersresponding to floods.133Big data initiatives such as that conductedby Orange in Cte dIvoire have shown that even a basic mobilephone traffic data set can enable conclusions about socialdivisions and segregation on the basis of ethnicity, language,religion or political persuasion. As Alex Pentland, director of the

Section 5


49/77

49/80


Human Dynamics Lab at MIT, points out, imagine what MuammarQaddafi would have done with this sort of data.134

Data integrity challenges also emerge where false positivesand false negatives may arise as systems are gamed or thewrong interpretations are applied to the data sets. The potentialfor fetishisation of data and the prospect that data will bemisinterpreted or manipulated to support particular viewpointsis high. As Steve Song points out,135the big and open datamovements are founded on an assumption that data, factsand truth are roughly equivalent. Data can be politicised ormisrepresented and yet come to represent an authoritative version

of the truth, having serious implications for decision-makingthat could deeply affect individuals life choices and futures.A pertinent example is that of nutrition policy in Ethiopia, wherea piece of data from a 2000 survey showing the high rate ofstunting in Amhara, a region that at that time was not listed asfood insecure, was used to show that malnutrition was a pervasiverather than acute problem and served as a motivating factor inthe formation of a national nutrition policy, despite this data beingincompatible with other pieces of data.136


50/77

50/80


Border surveillanceand security

The promise

The nexus between security and development and therecognition that security helps to create the necessaryconditions for development has long been at the heart ofmany development and humanitarian interventions.137The debate, however, about how to conceptualise and achievesecurity is an ongoing one, with the development community

generally moving towards referencing an understandingof human security over traditional conceptions of militarysecurity. Running in parallel with this discourse shift isthe increasing priority given the transfer of knowledge,tools, and technologies as a means of achieving securityin developing countries. Perceived as essential to ensuringthe effectiveness of humanitarian aid and the growth ofdemocratic institutions and the rule of law, foreign securityassistance and training takes the form of the transfer of

capacity, personnel and technologies to both the civilian andmilitary sectors. As noted by US President Obama in the 2010National Security Strategy:

Proactively investing in stronger societies and humanwelfare is far more effective and efficient than respondingafter state collapse. The United States must improveits capability to strengthen the security of states at riskof conflict

Date post:	24-Feb-2018
Category:	Documents
Upload:	divizioni
View:	229 times
Download:	0 times